04
Live Dashboards
// Nusantara Tech · HR Analytics · Tableau Public · Overview · Attrition · Performance · Compensation
Data Analyst · BI · Reporting · Data Operations · Remote
I’m Irsyad — I work across the full data workflow: cleaning and structuring messy data, writing SQL, building reports and dashboards, and surfacing the insights teams act on. Open to roles in data analysis, BI, reporting, and data operations.
I work at the intersection of data engineering, analysis, and communication. My day-to-day involves writing complex SQL against PostgreSQL, building analytical workflows in Python, and turning the output into dashboards that non-technical stakeholders can actually use.
My background spans HR & people analytics and retail & supply chain intelligence — building KPI frameworks, inventory health models, customer segmentation, and channel performance reports for audiences from team leads to General Managers.
I’m open to remote roles worldwide across data analysis, business intelligence, reporting, and data operations — from hands-on data cleaning and SQL modelling to the dashboards and recommendations leadership acts on.
Full analyst workflow on a ~4,500-employee dataset for fictional Nusantara Tech: messy raw CSV → validated clean data → multi-level SQL analysis → four-tab Tableau dashboard. Built to give the CHRO and People team a defensible view of attrition risk, performance distribution, and compensation equity — covering every stage from data pipeline to executive insight.
Handles mixed date formats, inconsistent categories, text-encoded numbers, duplicates, and missing values — with full referential-integrity validation at the end. Documented step-by-step in a notebook; packaged as a modular production script.
Querying five related tables in PostgreSQL: multi-table joins, CTEs, window functions (LAG, RANK), and a self-join on the manager hierarchy to model the org chart and surface attrition and performance patterns by department and level.
Four-tab Tableau Public story covering Overview, Attrition, Performance, and Compensation. Features a deliberate design system — gray+red highlight logic for anomalies, LOD calculated fields to resolve join fanout, and action filters linking scatter plots to bar chart breakdowns.
End-to-end analytics for Meridian Commerce Co. — a fictional omnichannel fashion & FMCG retailer across marketplace, D2C, and offline channels. Covers the full analyst stack: transaction data cleaning → SQL analysis → advanced customer analytics (RFM, cohort, A/B testing) → Looker Studio dashboard with geographic sales maps. Built to answer what a GM and Finance team act on every month.
Cleaning and validating five retail tables: SKU catalog, omnichannel transactions, inventory snapshots, traffic sessions, and marketing spend. Handles channel-name normalization, currency inconsistencies, negative stock entries, and duplicate transaction IDs.
Querying omnichannel transaction data in PostgreSQL: sell-through rate by SKU, ABC-XYZ inventory classification, aging stock flags (30/60/90 days), reorder point with lead time, and contribution margin by channel.
Customer segmentation using RFM scoring (Champions, At Risk, Lost, New). Retention cohort heatmaps by acquisition month. Statistical A/B testing of promo effectiveness — chi-square for conversion rates, t-test for AOV, with p-value and confidence intervals. Geographic sales distribution via choropleth maps.
// Nusantara Tech · HR Analytics · Tableau Public · Overview · Attrition · Performance · Compensation
A walkthrough of reading the query plan, finding the bottleneck, and rewriting for speed without changing a single output row.
Translating an org hierarchy into weighted KPIs — the data modeling decisions, and what leadership actually wanted to see.
When exact joins fail: using SequenceMatcher to standardize human-entered data, and how I validated the results.
Need someone reliable with your data — from cleanup to insight?
Let’s talk →